* fix(daemon): bound runtime --version probe so one wedged CLI can't block all runtimes
A CLI whose `--version` never returns (e.g. a brew-installed claude wedged
by a bun regression) stalled the daemon's sequential runtime registration
loop forever. Registration runs inside the blocking preflight that gates
/health, so the daemon never flipped from "starting" to "running" and every
runtime on the host appeared disconnected — not just the broken one.
detectCLIVersion now derives a 10s timeout context and sets cmd.WaitDelay so
a node/bun shim that leaves a child holding the stdout pipe open can't defeat
the timeout. A wedged probe now fails fast; the existing per-agent error skip
isolates the broken runtime and the rest register normally.
MUL-3812
Co-authored-by: multica-agent <github@multica.ai>
* test(agent): reap the hang script's orphaned child instead of leaking it
The MUL-3812 regression test spawned a background `sleep 60` that outlived
the killed parent and lingered for up to 60s on CI. The hang script now
records the child's PID and t.Cleanup reaps it, so no temporary process is
left behind.
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
When codex emits a single stdout line larger than the daemon's 10 MB
bufio.Scanner cap (typical trigger: thread/resume on a long-history
session), the reader goroutine returns scanner.Err()="token too long",
markProcessExited fails the in-flight RPC, and the lifecycle goroutine
enters its failure path. That path calls drainAndWait() — stdin.Close()
+ cmd.Wait() — before sending the failed Result. But cmd.Wait() never
returns: codex is alive and blocked writing the rest of the oversized
line into a stdout pipe nobody is reading, so it never reaches its
stdin read syscall and never sees the EOF. The lifecycle goroutine
therefore never sends Result{failed} to its caller, the outer daemon
blocks on the result channel, and the existing PriorSessionID-with-
empty-SessionID fallback never fires — the task is permanently
stalled and codex (Node wrapper + native Rust app-server) leaks until
the OS reaps them.
The cancel() that would have unblocked things via cmd.WaitDelay's
SIGKILL was registered as a defer AFTER drainAndWait, so LIFO defer
order put cancel last — drainAndWait blocks first, cancel never runs.
Fix:
1. drainAndWait now runs the existing graceful-then-cancel pattern
itself, in two bounded phases. Phase 1 waits for readerDone (capped
by codexGracefulShutdownTimeout, so we still give codex its OTEL
flush window on clean exits); on timeout it cancels the runCtx so
cmd.Cancel kills the tree and the reader unblocks. Phase 2 bounds
cmd.Wait() the same way for the scanner-overflow case, where
readerDone closed early but the process is still alive on a full
stdout pipe. The success-path cleanup that previously duplicated the
graceful-cancel pattern around readerDone collapses to a single
drainAndWait() call.
2. cmd.Cancel is set to send SIGKILL to the whole codex process group
(Setpgid via configureProcessGroup, signalProcessGroup on cancel)
instead of just the leader. This addresses YOMXXX's
orphaned-Codex-child concern: the Node wrapper and the native
app-server it spawns now both die when cleanup forces the kill,
rather than the native binary leaking as an orphan reparented to
init. configureProcessGroup is a no-op on Windows.
3. codexGracefulShutdownTimeoutNanos atomic.Int64 mirrors
opencodeTerminateGraceNanos so the regression test can shrink the
grace window from 10 s to 500 ms. Production code is unchanged
(default 10 s).
Outer daemon (daemon.go) already retries with a fresh session when
result.Status == "failed" && PriorSessionID != "" && result.SessionID
== ""; the failed Result now actually reaches it, so the recovery
fires on its own without any daemon-side change.
Tests:
- New regression TestCodexExecuteCleansUpWhenScannerOverflowsOnResume
spawns a fake codex that emits an 11 MB single-line thread/resume
response (trips the scanner cap) and then sleeps without re-reading
stdin. With the original drainAndWait body it blocks at the 10 s
executeFakeCodex deadline ("timeout waiting for result") — verified
by temporarily reverting just the helper body — and with the fix it
completes in ~1.3 s with Result.Status="failed",
Result.SessionID="" so the outer fallback can fire.
- Full codex test suite, full agent package, daemon + execenv +
repocache packages, go build ./..., and go vet on agent/daemon all
pass.
Out of scope (deferred to follow-up per YOMXXX): bumping the 10 MB
bufio.Scanner cap on codex / claude / copilot / cursor / hermes /
kimi / kiro / codebuddy / antigravity / qoder / openclaw / opencode
(pi already sits at 32 MB), and the shared bounded JSON-RPC line
reader that would eliminate the single-line-overflow risk class
entirely. Buffer size alone is not the fix — recovery behaviour is.
Refs: GH#4520
Co-authored-by: Eve <eve@multica-ai.local>
Co-authored-by: multica-agent <github@multica.ai>
On cancellation/timeout the opencode backend closed the stdout read end
immediately, leaving the child writing into a closed pipe. Every write then
returns EPIPE and, per anomalyco/opencode#33653, can spin an orphaned process
at 100% CPU — surfacing as high idle CPU after a cancelled task or daemon
restart (MUL-3655).
Cleanup now runs opencode in its own process group and, on cancel, drives a
graceful group-wide SIGTERM → grace → SIGKILL, closing the stdout pipe only as
a last-resort unblock once the tree has been signalled (SIGKILL is uncatchable,
so no member can write again — no EPIPE window). The group signal also reaps
tool subprocesses opencode spawned instead of orphaning them. WaitDelay remains
the hard backstop.
Adds unix tests covering the graceful path and the SIGTERM-ignored → SIGKILL
escalation, asserting the whole process group is reaped and the run never
deadlocks on the scanner. Windows behaviour is unchanged (no process groups).
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
agy's --print-timeout defaults to 5m when the flag is omitted, but the
daemon treated "omit the flag" as "no cap". In the default no-cap config
every Antigravity turn was therefore silently capped at 5 minutes: any
run whose build/tests outlived the budget had agy abort mid-turn, print
"Error: timed out waiting for response", and exit 0 — which the backend
recorded as a successful "completed" with truncated output (the reported
"Antigravity disconnects", MUL-3570 / #4453).
- Always pass --print-timeout: the configured cap when positive, else a
large value (24h) that defers to the daemon's idle/tool watchdogs.
- Detect agy's print-mode timeout marker in the run log and surface the
result as a timeout instead of a truncated success.
Verified by reproducing against agy 1.0.8 and with new unit + end-to-end
backend tests.
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
* feat(issues): unify run-enqueue decision behind WillEnqueueRun + preview endpoint
Collapse the issue update/batch enqueue copies into one service predicate
service.IssueService.WillEnqueueRun, shared verbatim with a new dry-run
endpoint POST /api/issues/preview-trigger so the four entry points stop
drifting (squad/self-loop/batch omissions, MUL-3375). The private-agent gate
stays at the HTTP boundary: write paths inject allow-all, preview injects the
real gate so it never leaks a private agent's readiness.
Add suppress_run to issue update/batch: the change applies but no run starts.
Remove the now-dead handler mirrors shouldEnqueueSquadLeaderOnAssign /
isSquadLeaderReady. service.Create and the comment trigger chain are untouched.
Tests: preview behavior, preview<->write-path match, batch aggregation,
member no-trigger, suppress_run skip, malformed-body 400.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* feat(issues): inject handoff note into assigned runs via first-class task field
Add an optional handoff_note carried by issue assign/promote into the run's
opening prompt and issue_context.md, via a dedicated agent_task_queue column
(migration 122) and a daemon assignment-handoff render branch — never a
fabricated comment, never trigger_comment_id (MUL-3375 §6.1).
Thread the note through enqueueIssueTask/enqueueMentionTask + WithHandoff
public variants and dispatchIssueRun; suppress_run or a parked write drops it
(no run = nothing to inject). Soft version gate: MinHandoffCLIVersion +
HandoffSupported, surfaced per-trigger as handoff_supported in the preview so
the UI can gray the note box on old daemons; the assignment never hard-fails.
Tests: daemon prompt + issue_context render via the assignment branch (not
quick-create/comment), version helper matrix, note persists on the task,
suppressed assign enqueues nothing.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* feat(issues): leave a display-only handoff record on the timeline
When an assign/promote with a handoff note starts a run, write one
type='handoff' timeline record via TaskService.RecordHandoff — a direct
Queries.CreateComment + timeline event that bypasses Handler.CreateComment, so
it never reaches triggerTasksForComment and cannot start a second run
(MUL-3375 §6.2, the must-not-retrigger invariant). Author is the actor who
handed off; body is the note. Migration 123 admits the 'handoff' comment type.
Recorded only on a real run start: suppress_run or a parked write writes
nothing. enqueueSquadLeaderTask now reports whether it enqueued so the trace
is gated on an actual dispatch.
Test: exactly one handoff record on assign-with-note, exactly one task (no
re-trigger), and no record when suppressed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* feat(issues): frontend plumbing for issue-trigger preview + handoff (core)
Add api.previewIssueTrigger + IssueTriggerPreviewSchema (zod parseWithFallback),
the use-issue-trigger-preview hook, issueKeys.issueTriggerPreview(+All) with WS
queue-state invalidation, suppress_run/handoff_note on UpdateIssueRequest, the
'handoff' CommentType, and stripping of the control fields from optimistic
update/batch cache patches (MUL-3375 §9).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* fix(issues): exclude handoff records from new-comment counting
type='handoff' is a display-only timeline record, not conversation. Exclude it
from CountNewCommentsSince so a handoff note never inflates the count of
"new comments to catch up on" fed to a claiming agent (MUL-3375 §12). Analytics
already excludes it (RecordHandoff is a direct write that emits no analytics
event), and the comment-trigger path is already bypassed.
Test: a handoff record does not bump the new-comment count; a real comment does.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* feat(issues): pre-trigger preview UI, handoff note, timeline card (web/desktop)
Wire the §9 frontend onto the preview endpoint + handoff fields:
- Delete the backlog blocking dialog (backlog-agent-hint*) and its modal type;
the over-eager nag is gone. Backlog awareness is now a passive label.
- RunConfirmModal: single assign + batch assign/status route here. Shows the
backend predicate's verdict ("将启动 @X" / "将启动 N 个" / parked), an optional
handoff note (assign only, soft-gated by handoff_supported), and 暂不启动 —
then applies via update/batch. No frontend guessing.
- create modal: passive CreateRunHint ("将启动 @X" / backlog parked).
- single status change stays a direct apply (unchanged).
- timeline: render type='handoff' as a distinct, non-interactive handoff card.
- i18n run_confirm + handoff_card across en/ja/ko/zh-Hans; drop backlog action
keys; locale parity green.
Tests: use-issue-actions (assign → run-confirm modal, member → direct),
create-issue + comment-card suites updated/green; views typecheck + lint clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* test(issues): use a valid anchor in the handoff count-exclusion test
CountNewCommentsSince filters id <> @anchor_id; SQL id <> NULL is NULL and
excludes every row, so an empty anchor made the control assertion read 0. The
production caller always passes a real anchor — mirror that with a non-matching
sentinel uuid.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* test(issues): RunConfirmModal apply logic (start/suppress/note-gate/batch)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* test(core): preview schema malformed/missing/null fallback coverage
Cover IssueTriggerPreviewSchema via parseWithFallback (MUL-3375): well-formed
parse, top-level + item default fills (empty/older backend), and fallback to
{ triggers: [], total_count: 0 } for malformed shapes, a dropped required
issue_id, a wrong-typed total_count, and null/non-object bodies — so the four
entry points degrade to "nothing will start" instead of throwing.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* refactor(issues): remove display-only handoff timeline record (留痕)
The handoff "留痕" timeline record (type='handoff' comment written on run
start) was judged superfluous and dropped per product call. This removes
only the display-only trace; the handoff NOTE injection into the run's
opening prompt + issue_context.md is untouched.
- backend: drop RecordHandoff + its call in dispatchIssueRun
- db: drop the `type <> 'handoff'` exclusion in CountNewCommentsSince and
migration 123 (comment_type_check reverts to the 4-type set from 001);
no production data exists for this unreleased feature
- frontend: drop the "handoff" CommentType, HandoffCard, and handoff_card
i18n (all locales)
- tests: drop handoff_count_test.go and the record-write assertions in
issue_trigger_preview_test.go (note-injection tests retained)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* feat(issues): dismissable run-confirm modal + team-handoff copy
Two fixes to the pre-trigger confirm modal (MUL-3375).
1. Dismissable: switch RunConfirmModal from AlertDialog to the standard
shadcn Dialog so it has the close (X) button + Esc + click-outside.
Previously the only choices were "start" / "don't start now" with no
way to abort the action entirely; dismissing now cancels with no write.
2. Copy: rework the action-surface wording away from the backend term
"run" toward team-handoff voice — 指派 / 开始 / 交接 (run stays only on
record surfaces). Unifies the note's three names to "交接说明", and
parallels the rewrite across en/ja/ko.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* chore(agent): bump handoff note min CLI version to 0.3.28
The daemon release that renders handoff notes ships in 0.3.28 (0.3.27
was the prior tag), so move the soft-gate threshold up. Below this the
note is silently dropped and the frontend grays the note box — assignment
is never blocked.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(issues): skip run-confirm when batch-moving issues to backlog
A move into backlog never starts a run (service/issue_trigger.go), so the
pre-trigger confirm modal degenerated to an empty "won't start" box with a
single Apply button — pure friction. Apply directly instead, matching the
single-issue status path. Other target statuses still route through the
modal.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(issues): refine pre-trigger preview hint and copy
- Move the create-issue run hint to a reveal band (grid 0fr→1fr) above the
property toolbar. It was sharing the footer button row and, lacking a
width constraint, reflowed the submit buttons whenever it appeared.
Restyle to a borderless, comment-style avatar+caption that is purely a
caption (non-interactive avatar).
- Distinguish squad from agent in the pre-trigger copy: a squad's leader
evaluates and delegates rather than "starting work" itself. Add
will_start_named_squad / will_start_squad / create_will_start_squad across
en/zh/ja/ko (reusing the squad_leader_* evaluate→arrange vocabulary) and
branch run-confirm + the create hint on squad assignees.
- Bold the assignee name in the run-confirm headline via a language-safe
sentinel split (no per-language prefix/suffix keys).
- Align zh "开始处理" → "开始工作" on the single-assign copy.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(issues): stub ActorAvatar in create-issue suite
CreateRunHint now renders an ActorAvatar for agent/squad assignees, which
pulls in getActorInitials/getActorAvatarUrl + the workspace/presence/navigation
hook tree. This form-focused suite only stubbed getActorName, so the
squad-forwarding test crashed with "getActorInitials is not a function". Stub
the avatar inert — its own behavior is covered elsewhere.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Walt <walt@multica.ai>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
* feat(agent): Qoder ACP runtime, chat reconnect recovery, and task linkage
- Add Qoder CLI backend (ACP transport, model discovery, blocked-args policy)
- Wire daemon/runtime config, docs, and UI provider assets
- Retry terminal task reports; add backoff unit tests
- Chat: SQL attach user message to task; handler + optimistic cache reconcile
- Invalidate chat/task-messages caches on WS reconnect; extract helper + tests
Co-authored-by: Orca <help@stably.ai>
Co-authored-by: Cursor <cursoragent@cursor.com>
* chore: drop non-Qoder changes (chat reconnect, task link, terminal report retries)
Keep only Qoder runtime, docs, daemon config/execenv, and UI provider assets.
Co-authored-by: Orca <help@stably.ai>
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(agent): harden Qoder ACP drain and wire project skills path
- Stop streaming to msgCh after reader wait so grace timeout cannot race close
- Resolve injected skills to .qoder/skills per Qoder CLI discovery
- Update AGENTS.md skill copy and add execenv tests
Co-authored-by: Orca <help@stably.ai>
Co-authored-by: Cursor <cursoragent@cursor.com>
* feat(qoder): add provider logo and wire MCP config into ACP sessions
- Add inline SVG QoderLogo component to provider-logo.tsx, replacing
the generic Monitor icon placeholder
- Add convertMcpConfigForACP helper to convert Claude-style MCP server
config (object map) into ACP array format for session/new and
session/resume
- Add unit tests for convertMcpConfigForACP covering stdio, SSE,
empty/nil, and multi-server cases
Co-authored-by: Orca <help@stably.ai>
* fix(test): capture both return values from InjectRuntimeConfig in Qoder test
Co-authored-by: Orca <help@stably.ai>
* fix(qoder): preserve remote MCP headers and promote provider errors
Addresses review feedback on #2461 (Bohan-J): two runtime-correctness
issues in the Qoder ACP backend.
1. Remote MCP headers were dropped. The bespoke convertMcpConfigForACP
only forwarded url/type, so an authenticated remote MCP server looked
configured in Multica but failed inside the Qoder session. Replace it
with the shared buildACPMcpServers helper (same path Hermes/Kimi/Kiro
use), which preserves headers as [{name, value}], sorts for
deterministic output, and handles remote transport aliases. Fail
closed on malformed mcp_config instead of silently dropping servers.
2. Provider failures could report as completed tasks. stderr was wired
via io.MultiWriter and the result was only promoted to failed when
output was empty, so a terminal upstream error (HTTP 429 / expired
token) racing a stopReason=end_turn with text still became
"completed". Switch to StderrPipe + an explicit copier, drain it
(bounded by the existing grace window, since qodercli can leave a
child holding the inherited fds) before the decision, and run the
shared promoteACPResultOnProviderError.
Tests: replace the convertMcpConfigForACP unit tests with two
end-to-end Qoder tests — one asserts the Authorization header reaches
the session/new payload as {name, value}, the other asserts a terminal
stderr error with non-empty output reports failed.
Co-authored-by: Orca <help@stably.ai>
* fix(qoder): align ACP session handling
Co-authored-by: Orca <help@stably.ai>
* fix(agent): guard qoder late output after drain
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: Orca <help@stably.ai>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
The daemon auto-grants Codex item/permissions/requestApproval requests by
echoing back the network / fileSystem profile scoped to the current turn.
Previously a malformed params payload and any permission key outside
network / fileSystem were dropped silently, so a future app-server
protocol that adds a new permission shape would be narrowed away with no
trace in daemon logs.
Log both cases (parse failure and dropped keys) without changing the
granted response. Addresses review nits on #4346 / MUL-3451.
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
* MUL-3284: add runtime_profile schema (custom runtime PR1)
Schema-only foundation for custom runtimes. Additive migration 120:
- New workspace-level `runtime_profile` table: the shared, team-visible
definition of a custom runtime (e.g. an in-house Codex wrapper).
protocol_family is CHECK-constrained to the exact backend list in
agent.New() (server/pkg/agent/agent.go). The only args column is
`fixed_args` (args every agent on the runtime must inherit); there is
deliberately no generic per-agent args field — those stay on
agent.custom_args.
- `agent_runtime.profile_id` (nullable, FK -> runtime_profile ON DELETE
CASCADE): NULL = built-in runtime, non-NULL = a registered instance of
a custom profile.
- Partial unique index agent_runtime_workspace_daemon_profile_key on
(workspace_id, daemon_id, profile_id) WHERE profile_id IS NOT NULL.
The legacy UNIQUE (workspace_id, daemon_id, provider) constraint is left
INTACT so the existing registration upsert
(ON CONFLICT (workspace_id, daemon_id, provider) in runtime.sql) keeps
resolving its arbiter and the server stays green. Converting that key to
a partial (WHERE profile_id IS NULL) index and making the upsert
profile-aware is PR2's registration work, not this migration.
Verified up + down against Postgres 17: full `migrate up` applies 120;
schema shows the table, column, partial index and intact legacy
constraint; functional checks pass (partial index blocks dup
(ws,daemon,profile), allows same profile on another daemon; CHECK and
display_name uniqueness reject bad input; legacy ON CONFLICT still
resolves; profile delete cascades to instances); down/up round-trip is
clean.
Co-authored-by: multica-agent <github@multica.ai>
* MUL-3284: drop DB FKs/cascade from runtime_profile migration (review fix)
Per review (house rule: no new database foreign keys / cascades; relational
integrity lives in the application layer):
- runtime_profile.workspace_id: drop REFERENCES workspace ON DELETE CASCADE
-> plain UUID NOT NULL.
- runtime_profile.created_by: drop REFERENCES "user" ON DELETE SET NULL
-> plain UUID.
- agent_runtime.profile_id: drop REFERENCES runtime_profile ON DELETE CASCADE
-> plain UUID.
CHECK constraints, UNIQUE (workspace_id, display_name), the workspace index,
and the partial unique index agent_runtime_workspace_daemon_profile_key are
unchanged. The legacy UNIQUE (workspace_id, daemon_id, provider) constraint
remains untouched.
Behavioral consequence: the database no longer auto-removes a profile's
agent_runtime instance rows on profile delete. That cleanup moves into PR2's
profile-delete path. Up-migration comments document this; down-migration
comment no longer references FKs/cascade.
Re-verified on Postgres 17: migrate up applies 120; no FK constraints exist on
the new columns; partial index still blocks dup (ws,daemon,profile_id); CHECK
and display_name uniqueness still reject bad input; deleting a profile now
leaves the runtime row orphaned (proving cascade is gone); down/up round-trip
clean with the legacy constraint intact.
Co-authored-by: multica-agent <github@multica.ai>
* MUL-3284 PR2 (server): runtime_profile CRUD + profile-aware registration
Server/DB half of the custom-runtime feature.
- Migration 121: convert the legacy UNIQUE (workspace_id, daemon_id, provider)
constraint on agent_runtime into a partial unique index scoped to built-in
rows (WHERE profile_id IS NULL). With 120's partial index on profile_id this
lets one daemon host the built-in provider AND custom profiles of the same
protocol family without collision.
- Queries: runtime_profile CRUD; ListEnabledRuntimeProfilesForWorkspace
(daemon-facing); CountAgentsByProfile + DeleteAgentRuntimesByProfile for the
app-layer cascade; profile-aware UpsertAgentRuntimeWithProfile; the built-in
UpsertAgentRuntime ON CONFLICT now spells out WHERE profile_id IS NULL so it
targets the right partial index. sqlc regenerated.
- agent.SupportedTypes / IsSupportedType: single-source protocol_family
whitelist, in lockstep with agent.New and the migration 120 CHECK.
- Handlers + routes: runtime_profile CRUD (member-read, admin-write) with
protocol_family whitelist validation, display_name uniqueness (409), and
fixed_args validation (no generic per-agent args — iron rule); a
daemon-token endpoint GET /api/daemon/workspaces/{id}/runtime-profiles;
DeleteRuntimeProfile does the app-layer cascade (delete instance rows then
profile, in one tx) and refuses (409) while active agents are bound.
- DaemonRegister accepts an optional per-runtime profile_id: validates the
profile belongs to the workspace and is enabled, registers via the
profile-aware upsert, and skips legacy hostname merge for custom rows.
AgentRuntimeResponse now carries profile_id.
Verified on Postgres 17: migrate up through 121; built-in + custom codex
coexist on one daemon; both upsert arbiters are idempotent; delete-by-profile
cascade removes only the custom instance; migrate down reverses 121 then 120
and replays clean. go build ./... and go vet pass; handler test package
compiles.
Daemon-side wiring (fetch profiles, PATH-resolve command_name, register with
profile_id, exec uses command_name) lands in a follow-up commit on this branch.
Co-authored-by: multica-agent <github@multica.ai>
* MUL-3284 PR2 (daemon): pull profiles, PATH-resolve, register, exec command
Daemon-side half of custom runtime profiles, against the server contract on
this branch.
- client.go: GetRuntimeProfiles(workspaceID) -> GET
/api/daemon/workspaces/{id}/runtime-profiles (mirrors GetWorkspaceRepos);
RuntimeProfile / RuntimeProfilesResponse types.
- types.go: Runtime gains profile_id (parsed from the register response so
runtimeIndex carries it).
- daemon.go:
* appendProfileRuntimes — called inside registerRuntimesForWorkspace before
the empty-runtimes guard. Best-effort fetch (older server 404s are logged
and swallowed; never fails registration). Per enabled profile: resolve
command_name via PATH (exec.LookPath, behind a `lookPath` test hook),
skip+log when absent, best-effort version probe, record the resolved
absolute path keyed by profile_id, and append a registration entry
{name, type=protocol_family, version, status:online, profile_id}. A
custom-only host (no built-in agents) still registers.
* profileCommandPaths map (guarded by d.mu) + recordProfileCommandPath /
customCommandPathForRuntime helpers.
* runTask: looks up the claimed task's RuntimeID -> profile command path and
overrides the executable path, synthesizing an AgentEntry so a custom
runtime runs even when the host has no built-in agent of the same
provider. provider (=protocol_family) is unchanged so agent.New still
selects the right backend.
- Tests: GetRuntimeProfiles request shape; profile runtime appended + path
recorded (custom-only host); profile skipped when command not on PATH;
profiles-fetch-404 is best-effort; customCommandPathForRuntime bookkeeping.
- agent: lockstep test pinning SupportedTypes to agent.New and the migration
120 protocol_family CHECK.
Iron rule honored: profile carries no generic per-agent args. fixed_args are
parsed and carried but intentionally NOT wired into the launch command yet
(optional/best-effort; explicit TODO(MUL-3284) in appendProfileRuntimes).
Verified: go build ./... clean; go vet ./internal/daemon/... clean;
go test ./internal/daemon/... pass (existing + 5 new); full
go test ./internal/handler/ suite passes against a migrated Postgres 17;
agent lockstep test passes.
Co-authored-by: multica-agent <github@multica.ai>
* MUL-3284 PR2: profile delete runs full archived-agent cascade (fix 500)
Review fix. DeleteRuntimeProfile previously guarded only on ACTIVE agents, but
agent.runtime_id is ON DELETE RESTRICT — a profile whose runtimes had only
ARCHIVED agents passed the guard, then DeleteAgentRuntimesByProfile hit the FK
and the handler 500'd.
Now it mirrors the mature runtime-delete cascade (DeleteAgentRuntime): in one
transaction it enumerates the profile's runtime rows, refuses (409) any with
active agents or active squads led by archived agents, then for each runtime
pauses autopilots pinned to its archived agents, drops archived squads led by
them, and hard-deletes the archived agents before removing the runtime rows
and the profile. No code path can now fall through to a raw FK error.
- queries: ListAgentRuntimeIDsByProfile (sqlc regen). Reuses the existing
per-runtime teardown queries (CountActiveSquadsWithArchivedLeadersByRuntime,
ListArchivedAgentIDsByRuntime, PauseAutopilotsByAgentAssignees,
DeleteSquadsByArchivedAgentsOnRuntime, DeleteArchivedAgentsByRuntime).
- tests: TestDeleteRuntimeProfile_ArchivedAgentCascade (archived-only profile
deletes cleanly: 204, runtime + archived agent + profile gone) and
TestDeleteRuntimeProfile_ActiveAgentBlocks (active agent → 409, survives).
Verified against Postgres 17: both new tests pass; full handler suite, daemon
tests, and agent lockstep test pass; go vet clean.
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: multica-agent <github@multica.ai>
Fixes Cursor agent token usage parsing for top-level camelCase, nested camelCase, and legacy nested snake_case result usage shapes. Includes tests for the locally verified nested camelCase stream-json output.
* feat(openclaw): support connecting to existing OpenClaw gateway (#3260)
When the daemon host is a lightweight dev machine or CI coordinator, the
heavy agent work (LLM inference, code execution, tool use) often belongs
on a more powerful remote server already running an OpenClaw gateway.
Multica historically hard-coded `openclaw agent --local`, forcing every
turn to execute in-process on the daemon host.
This change adds an opt-in gateway routing mode controlled per-agent via
`runtime_config`:
{
"mode": "gateway",
"gateway": { "host": "...", "port": 18789, "token": "...", "tls": false }
}
- Backend: ExecOptions gains OpenclawMode + OpenclawGateway; buildOpenclawArgs
drops `--local` when mode == "gateway". Per-task openclaw-config.json
wrapper pins gateway.{host,port,auth.{mode,token},tls} so users do not
need to edit the daemon host's `~/.openclaw/openclaw.json` to point at
a different endpoint.
- Daemon: AgentData carries the raw runtime_config; decoding is fail-soft
(malformed JSON falls back to local mode rather than blocking dispatch).
- API: gateway.token is masked to "***" on every GET; PATCH replays the
sentinel back, and the update handler restores the persisted token so
the round-trip never destroys the secret. Defense-in-depth masking on
WS broadcasts, plus String/MarshalJSON masking on the in-memory struct
to block stray `%+v` / json.Marshal leaks.
- UI: openclaw-only "Routing" tab on the agent detail page with mode
selector + structured endpoint form. Token uses a "saved — submit a
new value to rotate" UX and matching backend preserve hook.
Empty `runtime_config` keeps the historical embedded behaviour, so
existing agents are unaffected.
* fix(openclaw): address #3664 review — drop dead gateway field, gate pin on mode
Per Bohan-J's review:
- Remove the dead ExecOptions.OpenclawGateway field (+ its String/MarshalJSON and
the daemon.go construction block). It carried the plaintext bearer token but was
never read — buildOpenclawArgs only consumes OpenclawMode and the live gateway
path runs through execenv.OpenclawGatewayPin — so this narrows the secret's
footprint.
- Gate the gateway pin on mode=="gateway" in decodeOpenclawRuntimeConfig: a
{"mode":"local","gateway":{...,"token"}} payload no longer writes the token into
the 0o600 per-task wrapper that --local makes openclaw ignore.
- Warn on an unrecognized non-empty mode (e.g. "gatway") instead of silently
falling back to local.
- Run preserveMaskedGatewayToken in CreateAgent too, so a literal "***" at create
time can't persist as a real bearer token.
- Document the gateway host:port trust boundary (SSRF note for shared daemon hosts).
Adds regression tests for the local-mode pin drop and the unknown-mode warning.
* fix(agent): clear stale session id when a resumed ACP session is gone
When an agent's stored ACP session no longer exists on the runtime side,
session/resume still succeeds — hermes echoes the requested sessionId
back — so the failure only surfaces when session/prompt returns JSON-RPC
-32603 "Session not found". The backend then reported Status=failed with
the stale SessionID still set, which kept the daemon's resume-failure
fallback (gated on SessionID == "") from ever firing. The failed task
never updates the stored session, so every future mention on the same
(agent, issue) dispatched against the same dead id, forever (#4010).
handleResponse now returns a structured acpRPCError instead of a flat
string (rendered text unchanged), and the hermes/kimi/kiro prompt-error
paths clear the session id when the error is session-not-found class on
a resumed session. The daemon's existing retry then re-executes with a
fresh session and stores the replacement id, healing the mapping.
* fix(agent): clear stale session id when set_model hits a dead resumed session
With a model override, session/set_model runs before session/prompt,
so a resumed session that is gone on the agent side surfaces there
instead of at the prompt — and the error branch returned the stale
SessionID, so the daemon's fresh-session retry (gated on
SessionID == "") never fired. Apply the same clear-the-id fix in the
set_model error branch of all three backends.
Also relax isACPSessionNotFound to accept -32602: kimi-cli raises
RequestError.invalid_params({"session_id": "Session not found"}) for
every unknown-session path (src/kimi_cli/acp/server.py), so pinning
-32603 made the fix dead code for kimi. The wording gate keeps
unrelated invalid_params errors (e.g. "model not available") on the
preserve-the-id path.
Regression tests for all three backends: resumed session + model
override + set_model failing with each runtime's observed
session-not-found shape must yield status=failed with an empty
SessionID.
Treats Cursor's stream-json terminal `result` event as the protocol completion boundary so a lingering Cursor worker process can no longer hold the daemon task open after the agent has produced its final result.
- Tighten `cmd.WaitDelay` to 500ms (set before `Start()`)
- Set `resultSeen` and `cancel()` on terminal `result`
- Preserve completed/failed status across the cancellation via two `!resultSeen` guards in the post-loop status decision
- Add unix fake-CLI coverage for success and `is_error` terminal results
Codex telemetry was never reaching the OTLP collector for tasks run by the
daemon. The per-task config (including the [otel] block) is copied into
CODEX_HOME correctly, but the lifecycle goroutine closed stdin and then
immediately cancelled the run context, which SIGKILLs the app-server. Codex's
OTEL batch exporters only force-flush on a graceful shutdown, so the buffered
spans/metrics/logs were dropped before they could be exported — short tasks
lost everything, long tasks lost the final batch.
Let codex exit on its own after stdin EOF (running its shutdown + flush path)
and only force-cancel after a bounded grace period if it doesn't, so the reader
goroutine still can't block forever. Also set cmd.WaitDelay, matching the other
long-lived backends (claude, copilot, cursor, …).
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(daemon): wire agy --model and model discovery for Antigravity
agy 1.0.6 added a --model flag and an `agy models` catalog command, which
were the #1 blocker in the earlier agy-backend review (MUL-3125). The
antigravity backend already shipped but deliberately dropped opts.Model
because agy 1.0.1 had no way to select a model.
- buildAntigravityArgs now passes --model <display name> when opts.Model is
set; the value is the exact `agy models` display string (spaces + parens),
passed as a single exec arg so no shell quoting is needed.
- Block --model in custom_args so it can't override the managed value.
- ListModels("antigravity") enumerates via `agy models` (no static fallback:
agy silently no-ops on unrecognised models, so a stale guess would turn a
typo into a successful empty run).
- ModelSelectionSupported now returns true for every built-in provider; the
hook stays for any future model-less runtime.
- Daemon probe reads MULTICA_ANTIGRAVITY_MODEL for the daemon-wide default.
Co-authored-by: multica-agent <github@multica.ai>
* docs(providers): mark Antigravity model selection as supported
Antigravity gained --model in agy 1.0.6 (MUL-3125). Update the provider
matrix + prose (en/zh/ja/ko) from "managed internally / no --model" to
dynamic discovery via `agy models`, and refresh the now-stale picker
comments. Flag the display-string (not slug) shape and agy's silent no-op
on unrecognised values.
Co-authored-by: multica-agent <github@multica.ai>
* fix(daemon): reject unknown Antigravity model at spawn (MUL-3125)
agy exits 0 with empty output on an unrecognised --model, so a stale/typo'd
value would surface as a 'completed' but empty task. Validate opts.Model
against the `agy models` catalog in Execute before spawning: a non-empty
model the CLI does not advertise fails fast with an actionable error listing
the real choices. opts.Model is the single funnel for agent.model and the
MULTICA_ANTIGRAVITY_MODEL default, so this one check covers every source
(UI free-text, API, persisted value, env) — addressing Elon's review that a
UI-only guard is bypassable.
Validation is fail-OPEN: if the catalog can't be discovered we pass the
value through and let agy resolve it, so a discovery hiccup never blocks a
run. Pure antigravityModelError() is unit-tested (valid / unknown / near-miss
/ empty-model / empty-catalog); verified live against real agy 1.0.6.
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
Active long-running sessions are no longer killed by a fixed wall-clock deadline. Liveness is delegated to the idle watchdog (MULTICA_AGENT_IDLE_WATCHDOG, default 30m) with a larger in-flight-tool budget (MULTICA_AGENT_TOOL_WATCHDOG, default 2h). MULTICA_AGENT_TIMEOUT is an opt-in absolute cap (default 0 = no cap). The server-side 2.5h sweeper is unchanged as a coarse backstop.
Fixes#3745.
Raise pi and cursor model-list discovery timeouts 5s->15s to match opencode/ACP; openclaw stays 30s (sequential multi-spawn). Stop caching empty discovery results so a transient timeout doesn't keep the picker blank for the full TTL. Fixes#3729. MUL-2977.
* fix(agent): stop stripping user-facing CLAUDE_CODE_* config from child env
isFilteredChildEnvKey blanket-removed every CLAUDE_CODE_* var from the
spawned Claude Code child's environment. The intent was only to keep the
daemon's internal session markers from leaking, but CLAUDE_CODE_* is also
Anthropic's user-facing config namespace. On Windows this stripped the
user-set CLAUDE_CODE_GIT_BASH_PATH, so Claude Code could not locate
bash.exe, exited immediately, and every task failed with
"write claude input: write |1: The pipe has been ended."
Switch from prefixing the whole CLAUDE_CODE_ namespace to an exact-name
denylist of the internal runtime/session markers (CLAUDECODE,
CLAUDE_CODE_ENTRYPOINT, CLAUDE_CODE_EXECPATH, CLAUDE_CODE_SESSION_ID,
CLAUDE_CODE_TMPDIR, CLAUDE_CODE_SSE_PORT), still blanket-stripping the
wholly-internal CLAUDECODE_* namespace. Every other CLAUDE_CODE_* var
(GIT_BASH_PATH, USE_BEDROCK, USE_VERTEX, MAX_OUTPUT_TOKENS, ...) now
reaches the child. The internal-marker set was confirmed against the live
runtime, not guessed.
Fixes the whole class, not just git-bash: Bedrock/Vertex/etc. were
silently dropped the same way.
MUL-2940
Co-authored-by: multica-agent <github@multica.ai>
* fix(agent): keep CLAUDE_CODE_TMPDIR in child env
CLAUDE_CODE_TMPDIR is a documented, user-configurable temp-dir override
(public env-vars reference), not an internal per-session marker. Claude
Code creates its own per-session subdir under it, so inheriting it is
harmless — and stripping it would silently break a user's temp-dir
override the same way the broad prefix filter broke CLAUDE_CODE_GIT_BASH_PATH.
Drop it from the internal denylist (which now holds only the undocumented
per-process runtime markers: CLAUDECODE, CLAUDE_CODE_ENTRYPOINT,
CLAUDE_CODE_EXECPATH, CLAUDE_CODE_SESSION_ID, CLAUDE_CODE_SSE_PORT) and
assert it reaches the child.
MUL-2940
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
Parse the discovered catalog even when the model-discovery CLI exits non-zero (pi/opencode/cursor/openclaw) instead of discarding it and returning an empty model picker. Filter pi diagnostic lines so stale-pattern warnings don't coin bogus models. Fixes#3729. MUL-2977.
Newer opencode (1.15+) syncs its hosted free-model catalog over the
network on `opencode models`, which can take ~6s. The previous 5s cap
killed the command, discoverOpenCodeModels returned an empty list, and
the daemon reported it as a successful empty result — so the runtime
showed online but the model picker was empty ("暂无可用模型").
Fixes#3627
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
Adds OpenCode model variant discovery for thinking controls, passes saved thinking_level through opencode run --variant, and hardens verbose model parsing with fallback coverage.
Bug 1: detect copilot.cmd/.bat on Windows and invoke the sibling .ps1 directly via powershell -File, bypassing cmd.exe %* re-tokenisation that mangled the multi-line -p prompt. Shared rewriteCmdToPS1() now serves cursor, pi, and copilot.
Bug 2: filterCustomArgs (shared by all agent backends) strips one outer layer of shell quotes via unshellQuoteArg() before processing, so shell-style custom args like --deny-tool='write' no longer reach the CLI with literal quotes.
The claude backend wrote the full prompt to the child's stdin and closed
it before starting the stdout reader goroutine. With
--verbose --output-format stream-json the CLI emits a startup banner
before reading its first stdin frame; with no reader draining stdout, the
child blocks on its stdout write, never reads stdin, and our stdin Write
blocks until the per-task context fires. The field symptom is tasks
failing exactly at the 2 h per-task timeout with
"write |1: The pipe has been ended."
Move writeClaudeInput into its own goroutine so the prompt write and the
stdout drain proceed concurrently. Guard stdin close with sync.Once (it
can now be called from both the writer goroutine and, previously, the
result handler). Join the write result at cmd.Wait() and surface a write
failure as a "failed" status only when no result event arrived and no
session was established, so a genuine startup death still reports the
stderr tail.
Add a regression test that re-execs the test binary as a fake claude
which bursts 256 KiB to stdout before reading stdin, with a 128 KiB
prompt pushed at stdin — both past any plausible OS pipe buffer — so a
regression hangs until the test deadline instead of passing.
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Closes the runtime-side gap of #2106: previously `agent.mcp_config` was
honored only by Claude Code (via `--mcp-config <file>`); for OpenCode the
field was accepted by the API but silently ignored at execution time.
## Approach
OpenCode has no `--mcp-config` flag. Project the agent's `mcp_config`
into OpenCode via OPENCODE_CONFIG_CONTENT — OpenCode's general
inline-config injection environment variable, which accepts any subset
of OpenCode's config schema (model / agent / mode / plugin / mcp / …)
and merges at "local" scope after the project-config loop. MCP is the
only field this PR projects through that channel; if a future Multica
field needs the same channel it would assemble a combined config slice
before the env append.
The env-var route was deliberate. An earlier draft of this PR wrote
the translated MCP servers into <workdir>/opencode.json and removed
the file on cleanup; review (#3098) flagged that the task workdir is
reused across turns for the same (agent, issue), and any agent- or
user-written model / tools / permission settings in opencode.json
must survive across runs. OPENCODE_CONFIG_CONTENT avoids the workdir
entirely — nothing is written to disk, no cleanup is needed, and the
env entry dies with the spawned process.
OPENCODE_CONFIG_CONTENT was added to OpenCode in v1.4.10 (2025-09); the
official @opencode-ai/sdk uses the same env var to inject runtime
config, so the surface is stable. Verified empirically against
OpenCode 1.15.6 in our K8s runtime: `opencode debug config` returns
the injected mcp slice deep-merged with the user's global config,
and <workdir>/opencode.json is observably untouched.
## Translation surface
`agent.mcp_config` accepts two shapes for portability:
- Claude-style `{"mcpServers": {name: {url|command, ...}}}` is
translated into OpenCode's native form: `type: "local"|"remote"`,
`command` coerced to a string array, `env` renamed to `environment`.
- Native OpenCode `{"mcp": {name: ...}}` accepts the three shapes
OpenCode's schema permits and is strict-decoded against each:
- McpLocalConfig: `{type:"local", command:[…], environment?, enabled?, timeout?}`
- McpRemoteConfig: `{type:"remote", url:"…", headers?, oauth?, enabled?, timeout?}`
- bare override: `{enabled: bool}` (toggle a server inherited
from global / project config without redefining it)
Decoding uses `json.DisallowUnknownFields` so any field outside the
matching schema is rejected — matching OpenCode's
`additionalProperties: false`. Without this, a malformed payload
(e.g. `command: "node"` instead of `command: ["node"]`) would reach
OpenCode verbatim and either silently disable the server or crash
the CLI at startup.
Field-level checks the strict decoder doesn't catch:
- `timeout` must be a positive integer (rejects 0, negative, fractional)
- `oauth` must be either an object (validated against McpOAuthConfig)
or the literal `false`; primitives and `true` are rejected as ambiguous
- `oauth.callbackPort` must be in 1..65535 when set
## Precedence
Go's os/exec dedups `cmd.Env` by key keeping the LAST occurrence
(Go 1.9+). Appending OPENCODE_CONFIG_CONTENT after `buildEnv(b.cfg.Env)`
guarantees the daemon's value wins over any value the user happened
to put in `agent.custom_env` — which matches the intended semantics
(`mcp_config` is the authoritative daemon-managed field; `custom_env`
is the escape hatch). When that override happens we surface a warning
log so accidental clobbers are debuggable.
## Limitation (out of scope, accepted in review)
OpenCode also deep-merges its **global** config
(`~/.config/opencode/opencode.json`) into every session and exposes no
flag to disable that. Operators who want strict per-agent isolation
from the global layer can set:
```jsonc
// agent.custom_env on the platform
{ "XDG_CONFIG_HOME": "/tmp/opencode-isolated" }
```
…pointing at any directory without an `opencode/` subdir. OpenCode then
reads no global config and only honors what the daemon injects via
OPENCODE_CONFIG_CONTENT. Verified with `opencode debug config`.
## Changes
server/pkg/agent/opencode_mcp.go (new):
- buildOpenCodeMCPConfigContent — translates raw mcp_config into the
JSON string OpenCode accepts via OPENCODE_CONFIG_CONTENT, returns
"" when there's nothing to inject so the caller can skip the env
entry (avoids clobbering anything the user put in
agent.custom_env.OPENCODE_CONFIG_CONTENT)
- translateMCPConfigForOpenCode + helpers — Claude-style → OpenCode
native shape
- validateOpenCodeNativeMCPEntry + opencodeMCPLocal /
opencodeMCPRemote / opencodeMCPEnabledOnly / opencodeMCPOAuth
typed structs — strict-decode native-shape entries against the
schema (DisallowUnknownFields), plus targeted post-decode
assertions for timeout / oauth / callbackPort
server/pkg/agent/opencode.go:
- 12 lines of env injection in Execute(), placed AFTER buildEnv so
the daemon's value wins via os/exec dedup
- warning log when agent.custom_env duplicates the same key
- no on-disk state, no rollback closure, no post-run cleanup —
OPENCODE_CONFIG_CONTENT lives only in the spawned process env
server/pkg/agent/opencode_mcp_test.go (new):
- TestBuildOpenCodeMCPConfigContent_{Empty,Remote,Local,Native}
- TestBuildOpenCodeMCPConfigContent_NativeAcceptsAllSchemaFields —
covers each native variant round-tripping every optional field
(local with env+timeout+enabled; remote with headers+oauth-object+
timeout+enabled; remote with oauth: false; bare {enabled} override)
- TestBuildOpenCodeMCPConfigContent_RejectsMalformedNative — 31-case
table covering every constraint on Bohan-J's review: command must
be a string array, environment / headers values must be strings,
oauth must be an object or false, timeout must be a positive
integer, additionalProperties: false (per-shape allow-list checked
via DisallowUnknownFields)
- TestOpencodeBackendInjectsMCPConfigViaEnv — E2E happy path; fake
opencode binary captures $OPENCODE_CONFIG_CONTENT, asserts the
translated mcp slice is present AND <workdir>/opencode.json was
NOT written
- TestOpencodeBackendOmitsMCPEnvWhenEmpty — empty mcp_config does
NOT inject the env, preserving any value the user set in
agent.custom_env
- TestOpencodeBackendOverridesUserOpenCodeConfigContent — daemon
value wins via os/exec dedup keep-last
apps/docs/content/docs/providers.{en,zh}.mdx:
- flip OpenCode's MCP cell from ❌ to ✅
- reword the "MCP configuration: only Claude Code actually reads it"
section so OpenCode is included; describe each tool's mechanism
(Claude → `--mcp-config`, OpenCode → OPENCODE_CONFIG_CONTENT)
apps/docs/content/docs/install-agent-runtime.{en,zh}.mdx:
- update the Claude Code blurb (no longer "the only one")
- expand the OpenCode blurb to mention mcp_config support
- fix the now-broken /providers anchor
Refs #2106 (TS types and per-agent UI for mcp_config are separate
follow-ups, not in this PR).
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Claude Code now ships Opus 4.8 (claude-opus-4-8). Add it to the three
places that enumerate Claude models so the picker, thinking-level
catalog, and usage cost estimates all recognize it:
- claudeStaticModels(): list Claude Opus 4.8 (Sonnet 4.6 stays default)
- claudeModelEffortAllow: Opus supports the full low..max set incl. xhigh
- MODEL_PRICING: $5/$25 in, $0.50 cache read, $6.25 5m cache write —
same current-gen Opus tier as 4.5/4.6/4.7, confirmed against
platform.claude.com/docs/en/about-claude/pricing
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
* MUL-2764 feat(agent): wire mcp_config through ACP runtimes (Hermes / Kimi / Kiro)
The MCP config Tab (#3419) already lets admins save mcp_config on an
agent, and the daemon plumbs it through to `agent.ExecOptions.McpConfig`
for every runtime. Claude and Codex consume it; the three ACP runtimes
(Hermes / Kimi / Kiro) ignored the field and hardcoded an empty
`mcpServers: []` in their `session/new` requests.
Add `buildACPMcpServers` to translate the Claude-style `{"mcpServers":
{"<name>": {...}}}` object-of-objects into the array shape ACP requires
(`[{name, command, args, env: [{name,value}, ...]}, ...]` for stdio;
`[{type, name, url, headers: [...]}, ...]` for http/sse), then pass the
translated array on `session/new` (all three) and `session/load` (kiro
resume). Malformed JSON fails the launch closed — same contract Codex's
`renderCodexMcpServersBlock` uses — so users see a real error instead of
silently running with no MCP servers. Individual unclassifiable entries
(no command, no url) are skipped with a warning so one bad row can't
take MCP down for the rest of the agent.
Co-authored-by: multica-agent <github@multica.ai>
* MUL-2764 fix(agent): wire mcp_config through ACP resume + gate http/sse on capability
Addresses the two blockers Elon raised on #3439:
1. session/resume now carries mcpServers for Hermes and Kimi (Kiro's
session/load already did). Per the ACP Session Setup spec the resume
path re-attaches MCP servers, and without this a resumed task lost
access to MCP tools that a fresh task on the same agent would have
had. Pinned with new TestHermesResumeIncludesMcpServers and
TestKimiResumeIncludesMcpServers integration tests that inspect the
recorded wire request.
2. Added extractACPMcpCapabilities + filterACPMcpServersByCapability so
http/sse MCP entries get dropped (with a daemon-log warning naming
the entry) when the runtime's initialize response doesn't advertise
mcpCapabilities.http / .sse. Sending those entries to a stdio-only
runtime is a spec violation and reliably tanks session/new; now they
get filtered and the rest of the session still starts. Stdio entries
pass through unconditionally. Both backends wire the filter in right
after initialize so session/new and session/resume see the same
filtered list.
Also added TestKiroLoadIncludesMcpServersFromConfig — Elon flagged that
no test pinned "non-empty mcp_config actually reaches the wire" for
Kimi/Kiro, so the wire assertions go in for all three runtimes.
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
* feat(agent): add Antigravity runtime backend
Adds Google's Antigravity CLI (`agy`) as the 12th supported coding-tool
runtime, alongside Claude / Codex / Cursor / Copilot / Gemini / Hermes /
Kimi / Kiro / OpenCode / OpenClaw / Pi.
The CLI emits plain assistant text on stdout (no structured event
stream), so the backend streams stdout line-by-line as `MessageText`
events and accumulates the same text as the final `Result.Output`.
Session resumption uses `--conversation <id>`; because the conversation
UUID is not echoed on stdout, the daemon routes `--log-file` to a temp
file and recovers the id from the glog-formatted log lines.
MUL-2767
Co-authored-by: multica-agent <github@multica.ai>
* fix(agent): correct Antigravity capability contract from Elon review
- ModelSelectionSupported now returns false for antigravity. `agy` has no
--model flag and antigravityBackend deliberately drops opts.Model, so
the UI must render a disabled "Managed by runtime" picker instead of
an empty dropdown plus a silently-ignored manual-entry field. Also
stop seeding AgentEntry.Model from MULTICA_ANTIGRAVITY_MODEL — the
backend would silently ignore it.
- Antigravity skills now write to {workDir}/.agents/skills/, the CLI's
native workspace path (inherits Gemini CLI's layout per
https://antigravity.google/docs/gcli-migration). Previously they went
to the .agent_context/skills/ fallback that the CLI doesn't scan.
Runtime brief moves antigravity into the native-discovery branch and
local_skills.go points the user-level skill root at
~/.gemini/antigravity-cli/skills for Runtime → local skill import.
- Doc + UI comment sync: providers matrix / install-agent-runtime /
cloud-quickstart / agents-create / tasks (session-resume support) /
skills / README all now list Antigravity in the right buckets, and
the model-picker / model-dropdown comments cite antigravity (not the
stale hermes reference) as the supported=false example.
New tests: TestAntigravityModelSelectionUnsupported,
TestInjectRuntimeConfigAntigravity (native discovery wording),
TestWriteContextFilesAntigravityNativeSkills (.agents/skills/ landing,
.agent_context/skills/ NOT written).
Co-authored-by: multica-agent <github@multica.ai>
* feat(provider-logo): swap inline placeholder for real Antigravity PNG
Replaces the hand-drawn planet+arc placeholder with the official asset
shipped from Downloads. Stored next to the component; bundlers
(Next.js / electron-vite) resolve the PNG import to a URL string at
build time. Added a small assets.d.ts so packages/views' tsc accepts
PNG / SVG module imports — there was no prior asset usage in this
package to register the declaration.
---------
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
* MUL-2764: feat(agents): add MCP config tab to agent detail page
Backend already stores `mcp_config` and the daemon forwards it to the
runtime CLI via `--mcp-config`; this only adds the UI entry point.
The new tab presents a JSON editor that pretty-prints the existing
config, validates the buffer on every keystroke, and saves through the
existing `PUT /api/agents/{id}` path. Clearing the editor sends
`mcp_config: null`, which the handler reads as "wipe the column" and
the daemon falls back to the CLI's own default.
When the caller can't see secrets (agent actor, or a non-owner
non-admin member), the server already returns `mcp_config: null` with
`mcp_config_redacted: true`; the tab renders a read-only "configured
but hidden" state in that case so a non-privileged member cannot
silently overwrite an admin-owned config by saving an empty editor.
Co-authored-by: multica-agent <github@multica.ai>
* fix(agents): MCP tab — preserve in-flight edits + warn non-Claude runtimes
- Fix stale-editor sync: compare the local draft against the *previous*
original via a ref, so a background agent refetch updates an untouched
editor instead of being silently ignored. Without this, a draft equal to
the OLD original was treated as user-edited after the prop changed, and
the next Save would write the old config back over a concurrent admin
edit.
- Surface a notice inside the tab when the agent's runtime provider is not
Claude — today's daemon only forwards mcp_config via Claude's
--mcp-config, so saving on e.g. a Codex agent was silent but ineffective.
- Tests for both: rerender resyncs an untouched editor, rerender preserves
an in-flight edit, warning renders on non-Claude / hides on Claude.
MUL-2764
Co-authored-by: multica-agent <github@multica.ai>
* MUL-2764: feat(agents): codex MCP support + hide MCP tab on unsupported runtimes
- Backend: codex.go now translates agent.mcp_config (Claude-style
`{"mcpServers": {...}}`) into `-c mcp_servers.<name>=<inline-toml>`
flags for `codex app-server`, so MCP servers configured in the UI
reach Codex's per-task config layer. Bad mcp_config JSON downgrades
to a warn-and-skip so it can't break the agent launch.
- Frontend: AgentOverviewPane hides the MCP tab when the agent's
runtime provider doesn't read mcp_config — only `claude` and `codex`
are supported today, every other provider sees no MCP tab. The
previous in-tab warning is removed (no longer reachable).
- New shared helper `providerSupportsMcpConfig` lives in
`@multica/core/agents` so views and any future caller share one list
of MCP-aware providers.
- Tests: new go-side coverage for stdio + url + multi-server inputs,
TOML string escaping, malformed-input fallback, and arg ordering vs
custom_args; new views-side coverage for which providers surface the
MCP tab. En + zh-Hans copy and parity test refreshed.
Co-authored-by: multica-agent <github@multica.ai>
* MUL-2764: fix(agents): keep codex mcp_config secrets out of argv/logs
Move the agent's mcp_config from a `-c mcp_servers.<id>=<inline-toml>`
argv flag into a daemon-managed `[mcp_servers.*]` block inside the
per-task `$CODEX_HOME/config.toml`. mcp_servers.<id>.env is a documented
Codex config field and the UI already treats mcp_config as redacted for
non-admins; argv would have leaked those values into `ps aux` and the
`agent command` log line. The file is forced to 0600 to keep secrets in
the daemon owner's lane regardless of the seed file's mode.
Also drop user-supplied `-c/--config mcp_servers.*` entries from
custom_args. Codex `-c` is last-wins (verified against codex-cli 0.132.0),
so without filtering, a custom_args entry could silently shadow whatever
the MCP Tab saved.
Strip inherited `[mcp_servers.*]` tables from the per-task config.toml
when the agent has its own mcp_config, mirroring Claude's
`--strict-mcp-config`: avoids TOML "table already exists" errors on
name collisions and matches admin expectations that the MCP Tab is the
authoritative source for that task.
Co-authored-by: multica-agent <github@multica.ai>
* MUL-2764: fix(agents): codex mcp_config three-state semantics + custom_args compat
Address the third review pass:
1. Distinguish nil vs present-but-empty mcp_config. `{}` and
`{"mcpServers":{}}` now count as "admin saved an explicit (empty)
managed set" — strip inherited user `[mcp_servers.*]` and pin an
empty managed marker block. Only SQL NULL / JSON `null` map to
"absent" and fall back to the user's global `~/.codex/config.toml`.
This aligns Codex with the API's three-state contract (omit / null
/ object) and with Claude's `--strict-mcp-config` semantics.
2. Fail closed on `ensureCodexMcpConfig` errors and on managed
mcp_config without CODEX_HOME. Previous warn-and-launch would
silently inherit the user's global MCP servers and look identical
to a successful apply — exactly the surprise the MCP Tab is meant
to remove.
3. Only filter `-c mcp_servers.*` from `custom_args`/`extra_args`
when the agent has a managed mcp_config. Pre-MUL-2764 agents that
configured MCP via custom_args keep working; once an admin opts
in via the MCP Tab the daemon owns the `mcp_servers` namespace
and overrides are dropped (last-wins safety).
4. Update mcp_config locale intro to mention $CODEX_HOME/config.toml
instead of the now-removed `-c mcp_servers.*` argv path.
Tests:
- Split `TestEnsureCodexMcpConfigEmptyInputsAreNoop` into
`TestEnsureCodexMcpConfigAbsentLeavesUserTablesAlone` (nil/null)
and `TestEnsureCodexMcpConfigEmptyManagedSetStripsUserMcp` (`{}`,
`{"mcpServers":{}}`).
- Add `TestEnsureCodexMcpConfigEmptyManagedSetIdempotent` to pin
byte-identical reruns on the empty managed marker block.
- Add `TestHasManagedCodexMcpConfig` covering the eight relevant
inputs.
- Add `TestBuildCodexArgsPreservesCustomMcpOverridesWhenUnmanaged`
and `TestBuildCodexArgsDropsCustomMcpOverridesWhenManaged` to
pin the new gating.
- Add `TestCodexExecuteFailsClosedWhenMcpConfigInvalid` and
`TestCodexExecuteFailsClosedWhenManagedMcpButNoCodexHome` for the
Execute paths.
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
Pi is installed on Windows via npm, which lays down `pi.cmd` → `pi.ps1`
→ `node_modules/@mariozechner/pi-coding-agent/dist/cli.js`. The daemon
spawns Pi with `exec.Command("pi", ...)`; PATHEXT resolves that to
`pi.cmd`, and cmd.exe expands `%*` in the shim by re-tokenising the
original command line, which truncates any argv containing newlines.
buildPiArgs passes the full prompt as the last positional argv, so the
multi-line system+user prompt is silently cut at the first newline
before it reaches the JS entrypoint. The session JSONL then records
only the first line ("You are running as a chat assistant for a Multica
workspace.") and Pi replies as if the user message were missing
(GitHub multica-ai/multica#3306).
Mirror the existing cursor-agent fix: when LookPath resolves Pi to a
.cmd/.bat launcher and a sibling pi.ps1 exists, invoke PowerShell with
`-File <ps1>` directly and forward each arg as a discrete token. This
keeps us on the official launch path while skipping the cmd.exe %*
re-expansion. Falls back to the original launcher when pi.ps1 or
PowerShell can't be located.
The Windows test asserts the rewrite produces the expected argv and
that the multi-line positional prompt survives unchanged.
Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
The current cursor-agent CLI no longer has a 'chat' subcommand. The
positional 'chat' argument was silently treated as prompt text, leaking
into the user message (e.g. 'chat <actual prompt>').
Remove 'chat' from buildCursorArgs so the generated argv matches the
current cursor-agent CLI interface.
Fixes#3077
Fix Hermes ACP usage attribution to current model when agent.model is unset.
Also preserves cache-read token accounting and makes ACP model-list parsing more tolerant of snake_case payloads and Unknown display names.
* fix(agent): surface host OAuth token via env var on macOS isolation (MUL-2603)
Claude Code 2.x scopes the macOS keychain credentials entry by
sha256(CLAUDE_CONFIG_DIR)[:8], so the MUL-2603 isolation path strands
the child at "Not logged in" even after #3261 mirrored .claude.json:
the child looks up `Claude Code-credentials-<scratch-hash>`, the host
token is sitting in the no-suffix `Claude Code-credentials` entry.
Read the host OAuth token from the keychain via /usr/bin/security and
inject it as CLAUDE_CODE_OAUTH_TOKEN, which bypasses keychain lookup
entirely. Linux/Windows continue to use the .credentials.json mirror
(no-op there). Operator-pinned tokens and ANTHROPIC_API_KEY both take
precedence over the keychain reader.
Co-authored-by: multica-agent <github@multica.ai>
* fix(agent): tighten empty-value auth gate, pin Claude CLI env-scrub assumption (MUL-2603)
Empty-value gate
- `ANTHROPIC_API_KEY=` inherited from a login shell that conditionally
exports auth previously posed as an "operator pinned API-key auth"
choice and disabled the keychain reader, stranding the isolated child
at "Not logged in" even though no auth was actually selected.
- Custom_env `CLAUDE_CODE_OAUTH_TOKEN=""` (stale agent config) had the
same effect, plus would have shadowed a keychain-injected token in
libc env lookups that pick the first match.
- Both are now treated as noise: the empty entry is dropped from the
child env and the keychain reader runs unchanged. Two new unit tests
cover the os.Environ side (`...TreatsEmptyAnthropicAPIKeyAsUnpinned`,
`...HonorsNonEmptyAnthropicAPIKey`) and the custom_env side
(`...EmptyOAuthTokenInCustomEnvAsUnpinned`).
Env-scrub boundary
- Surfacing `CLAUDE_CODE_OAUTH_TOKEN` to the isolated child is only
safe because Claude Code itself drops that variable from the env it
hands to Bash / hook subprocesses, so a model-driven `printenv` can
never echo the secret into the agent transcript.
- Empirically verified against `claude` 2.1.121:
printf '...test -n "$CLAUDE_CODE_OAUTH_TOKEN" && echo SET || echo UNSET...' \
| CLAUDE_CODE_OAUTH_TOKEN=sk-canary-XYZ \
MUL2603_CONTROL=control-value \
claude --print --output-format text \
--allow-dangerously-skip-permissions --allowedTools Bash
returned `UNSET` for the OAuth token while the non-sensitive
`MUL2603_CONTROL` control returned `CONTROL-SET`, proving the CLI
scrubs only the auth env, not the env in general.
- Pinned this assumption in a new skip-gated regression test
(`TestClaudeCLIScrubsOAuthTokenFromBashSubprocess`) that boots the
real CLI with a canary token; failing the test means upstream
Claude Code stopped scrubbing and the passthrough must move off env
vars before MUL-2603 can ship.
Co-authored-by: multica-agent <github@multica.ai>
* fix(agent): gate keychain passthrough on default host dir, harden scrub test (MUL-2603)
Two follow-ups from the round-2 review on #3267:
1. Custom CLAUDE_CONFIG_DIR no longer pulls the default OAuth token.
Claude Code 2.x maps each config dir to its own suffixed
`Claude Code-credentials-<hash>` keychain entry, so an operator that
pins a managed/custom CLAUDE_CONFIG_DIR via custom_env or the
daemon-host env was getting the *daemon user's* default unsuffixed
entry injected into the isolated child — silently crossing accounts,
exactly the boundary mirrorHostClaudeJSONIfMissing already protects
for `.claude.json`. buildClaudeEnvWith now threads the effective
hostConfigDir through and only calls the reader when that dir is the
default `$HOME/.claude`. The new gate has a unit-level truth table
(TestIsDefaultHostClaudeConfigDir) plus a regression
(TestBuildClaudeEnvIsolatedSkipsKeychainForCustomHostConfigDir) that
makes a t.Fatal-armed reader prove the gate keeps the read off for
custom dirs.
2. Scrub e2e now asserts the control prong and the proof-of-execution
marker, not just "canary absent". The previous assertion would
false-pass on a model refusal, paraphrase, or "Bash gets no env at
all" upstream change. The strengthened version sets a non-secret
MUL2603_CONTROL alongside the canary OAuth token and asserts (a)
canary is NOT in the transcript, (b) CONTROL-SET IS in the
transcript (env propagation works for non-secrets — proves a
targeted scrub), (c) UNSET IS in the transcript (the Bash tool
actually ran AND saw the OAuth var as empty/unset). Code comment in
buildClaudeEnvWith and the test docstring now narrow the
security contract to the Bash tool subprocess only; hook subprocess
env-scrub is no longer claimed because it has not been verified.
Co-authored-by: multica-agent <github@multica.ai>
* test(agent): use per-run nonces in Claude scrub e2e to kill false-pass (MUL-2603)
Elon's round-3 review flagged that TestClaudeCLIScrubsOAuthTokenFromBashSubprocess
still false-passed: the proof markers "UNSET" / "CONTROL-SET" were literal
strings in the prompt, so strings.Contains matched them even when the model
only paraphrased the prompt without spawning Bash.
Replace the hard-coded markers with two per-run random hex nonces passed *only*
via env vars (MUL2603_UNSET_NONCE, MUL2603_CONTROL_NONCE). The prompt now
references the variable names, not the values, so the nonces can land in the
transcript only if a real Bash subprocess inherits the env vars and echoes
them. A paraphrasing or refusing model cannot fake nonces it never saw.
Also update the security-boundary comment in buildClaudeEnvWith to describe
the nonce-based proof.
Co-authored-by: multica-agent <github@multica.ai>
---------
Co-authored-by: multica-agent <github@multica.ai>
PR #3200 introduced per-agent `skills_local=ignore` isolation that
mirrors the host's Claude config dir into a per-task scratch dir,
omitting `skills/` to keep broken local skills out of the CLI's
discovery path. The mirror walks entries inside `hostConfigDir`
(default: `$HOME/.claude/`), but Claude Code's default layout stores
its main config — login state, project history — at
`$HOME/.claude.json`, a *sibling* of `~/.claude/` rather than inside
it. Once `CLAUDE_CONFIG_DIR=$ISOLATED` is set, the CLI looks for
`$ISOLATED/.claude.json`, finds only `backups/.claude.json.backup.*`
(those live inside `~/.claude/` and DO get mirrored), and exits with:
Claude configuration file not found at: …/.claude.json
Not logged in · Please run /login
— so every agent with `skills_local=ignore` on a host using the
default Claude layout dies on the first turn. Flipping the toggle back
to "merge" restores the host CLAUDE_CONFIG_DIR and recovers the agent;
that's the workaround Bohan flagged in MUL-2661.
Fix: after the existing `mirrorHostClaudeExceptSkills`, run a new
`mirrorHostClaudeJSONIfMissing` that pulls `$HOME/.claude.json` into
the scratch dir as `.claude.json` when (a) the dest doesn't already
have one and (b) the host source dir is the default `$HOME/.claude/`.
The custom-CLAUDE_CONFIG_DIR path is left alone because a pinned
custom dir is expected to be self-contained — silently borrowing
`$HOME/.claude.json` from a different account would mask credential
drift.
The helper goes through `createFileLink`, so it inherits the same
symlink → junction → hardlink → copy fallback chain the rest of the
mirror uses on Windows-without-Developer-Mode hosts.
Tests:
- `TestMirrorHostClaudeJSONIfMissing_DefaultLayoutMirrorsParentFile`
covers the happy path with an injected `homeDir`/`fileLink`.
- `TestMirrorHostClaudeJSONIfMissing_AlreadyPresentNoop` asserts a
pre-existing dest `.claude.json` (from a custom CLAUDE_CONFIG_DIR
mirror) is not overwritten.
- `TestMirrorHostClaudeJSONIfMissing_CustomHostDirSkipped` locks in
the custom-host-dir gate.
- `TestMirrorHostClaudeJSONIfMissing_MissingSourceNoop` documents the
env-var-auth-only / fresh-install case.
- `TestClaudeExecuteIsolatesProvidesClaudeJSONFromHome` is the
end-to-end MUL-2661 regression: a fake `\$HOME` with the default
split layout, `skills_local=ignore`, fake claude binary that prints
whatever `.claude.json` reaches the scratch dir. Asserts the file
rides through. Verified the test fails (with the documented
MUL-2661 error message) when the new mirror call is removed.
Verification:
- `go test ./pkg/agent/...` green (full agent suite).
- `GOOS=windows GOARCH=amd64 go vet ./pkg/agent/...` clean.
Co-authored-by: multica-agent <github@multica.ai>
